Validation of an ICD-9-based claims algorithm for identifying patients with chronic idiopathic/spontaneous urticaria.

BACKGROUND There is no specific International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for chronic idiopathic urticaria or spontaneous urticaria (CIU/CSU), a skin condition characterized by hives and angioedema lasting at least 6 weeks with no known cause. OBJECTIVE To validate an ICD-9-CM-based algorithm for identification of patients with CIU/CSU and thus facilitate claims-based research. METHODS Patient records were reviewed at 4 US practices. Patients included in the study were from a random sample of those identified by their physician as having CIU/CSU or because they met the following diagnosis-based algorithm: (1) at least 2 outpatient ICD-9-CM diagnosis codes 708.1, 708.8, or 708.9 at least 6 weeks apart or (2) 1 outpatient diagnosis of 708.1, 708.8, or 708.9 and 1 diagnosis of 995.1 at least 6 weeks apart. Data collected included ICD-9-CM codes, diagnoses of urticaria and allergy-related conditions, and medication use. Sensitivity and positive predictive value were calculated. The study was approved by the Western Institutional Review Board. RESULTS One hundred forty-nine patient records were reviewed (mean age 41.1 years; 73.8% were women; 69.1% were white): 115 were identified with the diagnosis-based algorithm, 90 were patients with "known CIU/CSU", and 56 were in the 2 groups. The mean duration of CIU/CSU was 2.9 to 3.1 years. The 2 cohorts most frequently had diagnoses of idiopathic urticaria, unspecified urticaria, and other specified urticaria. The diagnosis-based algorithm had a positive predictive value of 90.4% and a sensitivity of 71.1%. CONCLUSION The high positive predictive value suggests that patients identified using the algorithm are highly likely to have CIU/CSU. The 71.1% sensitivity suggests that most patients with CIU/CSU will be identified. The validation statistics support the use of the diagnosis-based algorithm in claims-based research, although future studies could refine the algorithm further.